On Transferability of grasp-affordances in data-driven grasping

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

174 Downloads (Pure)

Abstract

It has become a common practice to use simulation to generate large databases
of good grasps for grasp planning in robotics research. However, the existence of a generic simulation context that enables generation of high quality grasps that can be used in several different contexts such as bin-picking or picking objects from a table, has to our knowledge not been discussed in the literature. In this paper we investigate how well the quality of grasps simulated in a commonly used ”generic” context transfer to a specific context where the object is placed on a table. We generate a large database of grasp hypothesis for several objects, which we then evaluate in different dynamic simulation contexts eg. free float (no gravity, no obstacles), standing on table and lying on table. We present a comparison on the intersection of the grasp outcome space across the different contexts and quantitatively show that to generate reliable grasp databases, it is often required to use context specific simulation.
Original languageEnglish
Title of host publicationProceedings of the RAAD 2013
Number of pages9
Publication date11. Aug 2013
Publication statusPublished - 11. Aug 2013
Event22nd International Workshop on Robotics in Alpe-Adria-Danube Region - Portorož, Slovenia
Duration: 11. Sep 201313. Sep 2013

Workshop

Workshop22nd International Workshop on Robotics in Alpe-Adria-Danube Region
CountrySlovenia
CityPortorož
Period11/09/201313/09/2013

Fingerprint

Bins
Gravitation
Robotics
Planning
Computer simulation

Keywords

  • Robotic Grasping
  • Grasp-affordances
  • Dynamic Simulation
  • Data-driven grasping

Cite this

@inproceedings{ff310fa367dc4058948d47bb5cf8fa0d,
title = "On Transferability of grasp-affordances in data-driven grasping",
abstract = "It has become a common practice to use simulation to generate large databasesof good grasps for grasp planning in robotics research. However, the existence of a generic simulation context that enables generation of high quality grasps that can be used in several different contexts such as bin-picking or picking objects from a table, has to our knowledge not been discussed in the literature. In this paper we investigate how well the quality of grasps simulated in a commonly used ”generic” context transfer to a specific context where the object is placed on a table. We generate a large database of grasp hypothesis for several objects, which we then evaluate in different dynamic simulation contexts eg. free float (no gravity, no obstacles), standing on table and lying on table. We present a comparison on the intersection of the grasp outcome space across the different contexts and quantitatively show that to generate reliable grasp databases, it is often required to use context specific simulation.",
keywords = "Robotic Grasping, Grasp-affordances, Dynamic Simulation, Data-driven grasping",
author = "Rytz, {Jimmy Alison} and Lars-Peter Ellekilde and Dirk Kraft and Petersen, {Henrik Gordon} and Norbert Kr{\"u}ger",
year = "2013",
month = "8",
day = "11",
language = "English",
booktitle = "Proceedings of the RAAD 2013",

}

Rytz, JA, Ellekilde, L-P, Kraft, D, Petersen, HG & Krüger, N 2013, On Transferability of grasp-affordances in data-driven grasping. in Proceedings of the RAAD 2013. 22nd International Workshop on Robotics in Alpe-Adria-Danube Region, Portorož, Slovenia, 11/09/2013.

On Transferability of grasp-affordances in data-driven grasping. / Rytz, Jimmy Alison; Ellekilde, Lars-Peter; Kraft, Dirk; Petersen, Henrik Gordon; Krüger, Norbert.

Proceedings of the RAAD 2013. 2013.

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

TY - GEN

T1 - On Transferability of grasp-affordances in data-driven grasping

AU - Rytz, Jimmy Alison

AU - Ellekilde, Lars-Peter

AU - Kraft, Dirk

AU - Petersen, Henrik Gordon

AU - Krüger, Norbert

PY - 2013/8/11

Y1 - 2013/8/11

N2 - It has become a common practice to use simulation to generate large databasesof good grasps for grasp planning in robotics research. However, the existence of a generic simulation context that enables generation of high quality grasps that can be used in several different contexts such as bin-picking or picking objects from a table, has to our knowledge not been discussed in the literature. In this paper we investigate how well the quality of grasps simulated in a commonly used ”generic” context transfer to a specific context where the object is placed on a table. We generate a large database of grasp hypothesis for several objects, which we then evaluate in different dynamic simulation contexts eg. free float (no gravity, no obstacles), standing on table and lying on table. We present a comparison on the intersection of the grasp outcome space across the different contexts and quantitatively show that to generate reliable grasp databases, it is often required to use context specific simulation.

AB - It has become a common practice to use simulation to generate large databasesof good grasps for grasp planning in robotics research. However, the existence of a generic simulation context that enables generation of high quality grasps that can be used in several different contexts such as bin-picking or picking objects from a table, has to our knowledge not been discussed in the literature. In this paper we investigate how well the quality of grasps simulated in a commonly used ”generic” context transfer to a specific context where the object is placed on a table. We generate a large database of grasp hypothesis for several objects, which we then evaluate in different dynamic simulation contexts eg. free float (no gravity, no obstacles), standing on table and lying on table. We present a comparison on the intersection of the grasp outcome space across the different contexts and quantitatively show that to generate reliable grasp databases, it is often required to use context specific simulation.

KW - Robotic Grasping

KW - Grasp-affordances

KW - Dynamic Simulation

KW - Data-driven grasping

M3 - Article in proceedings

BT - Proceedings of the RAAD 2013

ER -